Hysteresis of Soil-water Characteristic Curve for Soil with Bimodal Grain-Size Distribution

Abstract

Many regions are characterized by residual soils which are often gap-graded and are associated with dual porosity and a bimodal grain size distribution. Gap-graded soils also exhibit bimodal unsaturated hydraulic properties, therefore, its interaction with moisture is unique. To effectively address geotechnical challenges associated with these soils, it is essential that their geotechnical analyses incorporate the soil properties, to realistically represent actual soil behavior and field conditions. Changes in soil moisture during drying (from evaporation) and wetting (from precipitation and snowmelt) and pore suction generally affect the overall soil's behavior. The most important soil parameter that illustrates the link between the water content in the soil pores and the associated soil suction is the Soil-Water Characteristic Curve (SWCC). The soil moisture conditions throughout the drying and wetting cycles are represented by hysteresis. The hysteresis phenomenon refers to the different paths the boundaries drying and wetting curves of the SWCC follows during drying and wetting process of a soil, where the drying curve always shows higher water content more than the wetting curve, at the same suction. The wetting curve is more important for practical applications than the drying curve in geotechnical engineering because structural failures, particularly in soils with dual porosity (bimodal soils), typically occur during the wetting phase. However, because measuring SWCC during wetting is expensive, time-consuming, and challenging, the wetting curve has frequently been ignored. Despite the importance of the wetting curve in practical applications and the availability of gap-graded soil in many places across the world, before this work there is no mathematical model developed to predict the SWCC during the wetting process to solve the mentioned measurement challenges. Therefore, analyses and designs are mostly conducted using only the drying curve, entirely ignoring the SWCC hysteresis. But analyses based on the drying curve alone do not accurately represent actual soil conditions, potentially leading to flawed designs and structural failures. This study introduced a new mathematical model for estimating the wetting soil-water characteristic curve (SWCC) in bimodal soils. The model uses parameters with clear physical meaning tied to wetting-SWCC related variables, making it highly useful for practical engineering applications and numerical simulations. To validate the model, hysteretic SWCCs for multiple soil types were measured in the laboratory. In addition, an alternative equation was developed to fit discrete SWCC data obtained in the laboratory. Model performance was assessed using the coefficient of determination (R²) and the Root Mean Square Error (RMSE). The results demonstrated excellent accuracy, with an average R² of 0.9945 and an average RMSE of 0.0263 across all soil samples. Given the critical influence of moisture-induced failures in geotechnical engineering, the proposed models provide valuable tools for improving analysis and design. To highlight the importance of using actual soil properties and moisture conditions, numerical slope stability analyses were equally conducted for three scenarios: using only drying curves, only wetting curves, and full SWCC hysteresis. The results show that simulations incorporating full hysteresis more closely reflect real field conditions involving cyclic wetting and drying.